CONTENTS

1. CordellGleditschKernSaavedra-Lux-MeasuringInstitutionalVariation.R = R code used to produce all tables, figures and descriptive statistics in manuscript

2. constitutions.csv = 97 American Indian constitutions
- doc.id = unique document id number from our database of AIAN constitutional documents                       
- doc.name = name of the text file (doc.id.txt)                    
- tribe = AIAN nation name                       
- year = year that the constitution was approved                        
- decade = decade that the constitution was approved                    - const.clean.text = original text of the entire constitution, converted to lower case with unnecessary white space, punctuation, encoded text and non-ascii characters removed            
- judiciary.sect.clean.text = original text of the judicial section, converted to lower case with unnecessary white space, punctuation, encoded text and non-ascii characters removed     
- const.dict.text = const.clean.text, stemmed words, used to create the dictionary-based variables              
- judiciary.sect.dict.text = judiciary.sect.clean.text, stemmed words, used to create the dictionary-based variables    
- const.sml.text = const.dict.text, with numbers and stop words removed, used to create the supervised machine learning-based variables              - judiciary.sect.sml.text = judiciary.sect.dict.text, with numbers and stop words removed, used to create the supervised machine learning-based variables     
- const.overlap.ck = constitutions in our sample that overlap with Cornell & Kalt’s (2000) data (1 = yes, 0 = no)            
- const.words = total number of words included in the constitution                 
- judiciary.sect.words = total number of words included in the judicial section        
- judiciary.dict.words = number of times a term in our judiciary dictionary appears in the constitution       - independ.dict.words = number of times a term in our judicial independence dictionary appears in the judicial section         
- judiciary.ck = JUDICIARY variable from Cornell & Kalt (2000)                
- independ.ck = INDEPEND variable from Cornell & Kalt (2000)               
- judiciary.repl.dict = JUDICIARY replication variable, using a dictionary approach         
- independ.repl.dict = INDEPEND replication variable, using a dictionary approach         - judiciary.repl.sml.logistic = JUDICIARY replication variable, using a logistic regression supervised machine learning model  
- judiciary.repl.sml.nb = JUDICIARY replication variable, using a naive bayes supervised machine learning model       
- judiciary.repl.sml.svm = JUDICIARY replication variable, using a support vector machine supervised machine learning model      
- independ.repl.sml.logistic = INDEPEND replication variable, using a logistic regression supervised machine learning model   
- independ.repl.sml.nb = INDEPEND replication variable, using a naive bayes supervised machine learning model       - independ.repl.sml.svm = INDEPEND replication variable, using a support vector machine supervised machine learning model       